ICML 2026 spotlight！这帮人让多智能体在潜在空间用思想沟通，不用说话，比传统文本交互快4倍，准确率还高14.6%。
LatentMAS 论文已被 ICML 2026 接收为 spotlight 展示。该方法让 LLM 智能体直接通过隐藏嵌入进行推理和通信，无需文本解码或额外训练。在复杂推理任务上准确率提升最高达 14.6%，推理速度提高 4-4.6 倍，输出 token 使用减少 70.8%-83.7%。采用自回归潜在思维、KV-cache 传输等机制实现无训练协作。该技术可即插即用于现有 LLM，推动多智能体系统从文本交流转向潜在空间协同思考。
源：https://t.co/F9WflQaL7n
源： x.com/Jiaru_Zou/stat… Jiaru "Rubin" Zou @Jiaru_Zou Excited to share that #LatentMAS has been accepted to ICML 2026 as a spotlight! 💻Code: github.com/Gen-Verse/Late… Y 📄Paper arxiv.org/abs/2511.20639 ic We push multi-agent collaboration into the latent space — beyond human language. Most multi-agent systems rely on text: agents reason in words, exchange messages, and repeatedly decode/re-encode information. But language can be slow, lossy, and unnecessarily constrained. 💡LatentMAS takes a different path: LLM agents reason and communicate directly through hidden embeddings. No text decoding. No extra training. No token-level message passing. Instead, agents collaborate through: 🧠 Autoregressive Latent Thoughts — hidden-state-level reasoning steps 🔁 Latent Communication — information sharing via KV-cache transfer 📌 Input-output Alignment — keeping latent representations in-distribution 🚀 Training-free Collaboration — plug-and-play with existing LLMs Why it matters: ✅ Up to +14.6% better accuracy on complex reasoning tasks ⚡ 4-4.6x faster end-to-end inference ✂️ 70.8%–83.7% reduction in output token usage A step toward multi-agent systems that collaborate not by speaking more, but by thinking together in latent s #MultiAgentSystems S #ModelCollaboration o #LatentReasoning a #LLM n #AgenticAI e #ICML I #ICML 🔗 View Quoted Tweet 💬 0 🔄 0 ❤️ 0 👀 184 ⚡